Materials Science and Engineering C 33 (2013) 49524957

Contents lists available at ScienceDirect

Materials Science and Engineering C

journal homepage: www.elsevier.com/locate/msec

Characterisation of micro-sized and nano-sized tungsten oxide-epoxy

composites for radiation shielding of diagnostic X-rays
N.Z. Noor Azman a,b, S.A. Siddiqui a, I.M. Low a,
a
b

Department of Imaging and Applied Physics, Curtin University, GPO Box U1987, Perth, WA 6845 Australia
School of Physics, Universiti Sains Malaysia, 11800 Penang, Malaysia

a r t i c l e

i n f o

Article history:
Received 6 May 2013
Received in revised form 5 August 2013
Accepted 18 August 2013
Available online 28 August 2013
Keywords:
X-ray shielding
Micro-sized WO3-epoxy composites
Nano-sized WO3-epoxy composites
X-ray transmission
Filler loading

a b s t r a c t
Characteristics of X-ray transmissions were investigated for epoxy composites lled with 210 vol% WO3 loadings using synchrotron X-ray absorption spectroscopy (XAS) at 1040 keV. The results obtained were used to determine the equivalent X-ray energies for the operating X-ray tube voltages of mammography and radiology
machines. The results conrmed the superior attenuation ability of nano-sized WO3-epoxy composites in the energy range of 1025 keV when compared to their micro-sized counterparts. However, at higher synchrotron radiation energies (i.e., 3040 keV), the X-ray transmission characteristics were similar with no apparent size effect
for both nano-sized and micro-sized WO3-epoxy composites. The equivalent X-ray energies for the operating Xray tube voltages of the mammography unit (2549 kV) were in the range of 1525 keV. Similarly, for a radiology unit operating at 4060 kV, the equivalent energy range was 2540 keV, and for operating voltages greater
than 60 kV (i.e., 70100 kV), the equivalent energy was in excess of 40 keV. The mechanical properties of epoxy
composites increased initially with an increase in the ller loading but a further increase in the WO3 loading
resulted in deterioration of exural strength, modulus and hardness.
Crown Copyright 2013 Published by Elsevier B.V. All rights reserved.

1. Introduction
Hitherto, numerous analytical methods have been developed to investigate the effect of the particle size of a material on the X-ray attenuation for various incoming X-ray energies including scattered gammarays and X-rays [111]. It is widely believed that nano-sized particles
are able to disperse more uniformly within the matrix with less agglomerations when compared to micro-sized particles, thus improving the Xray attenuation ability of the material [6,12,13]. For instance, Hoyska
(1969) found that the intensity of scattered radiations increased with
increases in the grain size of a material. This size effect has been observed in a sand matrix and for samples containing heavy elements
such as iron or barium [4].
Filler-reinforced polymers have gained increasing attention from
X-ray technologists in radiation shielding since polymers have great
potential in many important applications by virtue of their unique
properties, such as low density, the ability to form intricate shapes,
optical transparency, low manufacturing cost and toughness. One of
the ller-reinforced polymers commonly used for radiation shielding
is lead acrylic [1416]. Moreover, some researchers have also tried to
synthesise nano-sized ller-reinforced polymers for radiation shielding
by virtue of the size effect in X-ray attenuation [2,6,17]. For instance,
a recent study by Botelho et al. [13] revealed that the attenuation for

Corresponding author. Tel.: +61 8 9266 7544; fax: +61 8 9266 2377.
E-mail address: j.low@curtin.edu.au (I.M. Low).

X-ray beams generated from low tube voltages (i.e., 2630 kV) in
nanostructured copper oxide (CuO) was better than microstructured
CuO. However, no signicant difference in attenuation was observed
for X-rays generated from higher tube voltages (i.e., 60102 kV). A
similar conclusion on this size effect in X-ray attenuation was made
by Kunzel et al. [18] for a nanostructured CuO-polymer system.
In a recent work on WO3-lled epoxy composites [19], we investigated the effect of nano-sized and micro-sized WO3 ller-epoxy
composites on X-ray attenuation in the X-ray tube voltage range of
22127 kV generated by a mammography unit and a general radiography unit. The equivalent X-ray energies for the various X-ray tube
voltages used were in the range of 17.560 keV, which conformed to
our expectation since the equivalent X-ray energies for a mammography unit were 17.5, 19.6, 20.2 and 22.7 keV, which are the characteristic energies of molybdenum and rhodium, while the equivalent
X-ray energy for a radiology unit is about one-third of the X-ray tube
voltage used. The results showed that nano-sized WO3 was more effective
than micro-sized WO3 in X-ray attenuation only in the low X-ray
tube voltage range of 2235 kV, but this size effect was not apparent
at the higher X-ray operating tube voltage range of 40120 kV [19].
Hence, the aim of this work was to verify our previous work on
X-ray transmission in WO3-lled epoxy composites by using synchrotron
radiations as the X-ray source for the characteristic (monochromatic) Xray energy range of 1040 keV. The results obtained were compared
with those of previous work [19] to determine the equivalent energy
range of the previous machines used (a mammography unit and a
radiology unit).

0928-4931/$ see front matter. Crown Copyright 2013 Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.msec.2013.08.023

N.Z. Noor Azman et al. / Materials Science and Engineering C 33 (2013) 49524957

4953

2. Experimental procedure

2.4. Flexural tests

2.1. Sample preparation

Three-point bending tests were used to determine the exural

strength and modulus of the WO3-epoxy composites containing different loadings of nano-sized and micro-sized WO3. Specimens with
dimensions 60 mm 10 mm 2 mm were prepared for the test
according to the ASTM D790-03 standard [20] on a universal testing
machine (LLOYD Instruments). A minimum of three samples was
tested for each composite and the average results were taken. In
these measurements, the samples were tested to the applied load
with the results calculated by the NEXYGEN Plus software.

Nano-sized (b 100 nm) tungsten oxide and micro-sized (~20 m)

tungsten oxide (WO3) were used as the ller for synthesising WO3epoxy composites. The former was obtained from Sigma-Aldrich and
the latter (FR251) from Fibreglass and Resin Sales. The particle sizes
were provided by Sigma-Aldrich.
Details about the preparation of the WO3-epoxy composite samples
are available in [19]. For the current work, the thickness of the samples
was set at 2.0 mm. The list of the prepared samples with the different
volume percentages of WO3 is shown in Table 1.
2.2. Measurements of X-ray transmission
This work was done using the X-ray absorption spectroscopy (XAS)
beamline at the Australian Synchrotron located in Melbourne. Experiments were carried out in the energy range of 1040 keV using an Si
(311) monochromator and a beam size on the sample of about 0.25
0.25 mm2 and a photon ux of about 1011 ph/s. Data on X-ray transmission (I/Io) were collected in an ionisation chamber after the sample and
compared to the incident beam ux. For each sample, 20 readings were
recorded at each energy level. To normalise the data collected, readings
were also recorded for an empty sample holder. The average I/Io was
calculated and plotted as a function of ller loadings (WO3 vol%) for
each synchrotron radiation energy. The X-ray transmission (I/Io) was
related to the linear attenuation coefcient () and the thickness of
the samples (t) through Eq. (1):
I=I o exp t

Next, all the samples were examined again with a mammography

unit and a radiology unit according to previous experiments [19].
2.3. Powder diffraction
Powder diffraction (PD) measurements were conducted to identify
the crystallite size of the WO3. This was done at the Australian Synchrotron in Melbourne, on the Powder Diffraction beamline. Mythen detectors were used to record the diffraction patterns in the 2 range
of 060 at a xed wavelength of 1.13 . The crystallite size (L) of
the nano-sized WO3 was determined using the Scherrer equation
as follows (Eq. (2)):
L

k
FWHM cos

2.5. Indentation hardness test

The hardness values of polished WO3-epoxy composites were determined using a Rockwell hardness tester with scale H. The hardness
measurements were conducted using an indentation load of 588.4 N
and ball diameter of 0.3 cm. Five measurements were conducted for
each sample in order to obtain an average value.
3. Results and discussion
3.1. Characteristics of X-ray transmissions
Since the synchrotron radiations of the XAS beam line contain a large
range of energies (i.e., 1040 keV), the thickness of the samples was set
at 2 mm to ensure that the detector was able to obtain a meaningful Xray transmission reading for the lower energy range without being totally absorbed by the samples. As shown in Fig. 1, there was an obvious
difference in X-ray transmissions between the micro-sized WO3-epoxy
and nano-sized WO3-epoxy composites of the same WO3 vol% at the
energy range of 1020 keV. With a further increase of synchrotron
energy to greater than 20 keV, there was no difference in X-ray transmissions between these two composites, thus indicating the absence
of size effect at play. The results show that for all the WO3 loadings in
an epoxy matrix, the ratio of the X-ray transmission of the micro-sized
WO3-epoxy composite (I/Io)m relative to the nano-sized WO3-epoxy
composite (I/Io)n, (I/Io)m/(I/Io)n remained at ~1.0 for the synchrotron
energy range of 2540 keV. On the other hand, the ratio (I/Io)m/(I/Io)n
was 1.152.3 for all WO3 loadings (see Fig. 2). The values determined
for (I/Io)m/(I/Io)n at these energy ranges indicate that the nano-sized
WO3-epoxy samples absorbed more low energy X-rays than their
microsized WO3-epoxy counterparts.
Further investigations were conducted to verify our previous results
[19] obtained from a mammography unit and a radiography unit. In this
investigation, all the same measurements from the previous work were
repeated with samples of 2 mm in thickness. Since a mammography

where k is a constant depending upon the crystal shape and size

(k = 0.90), is the wavelength, FWHM is the full-width half maximum of the peak and 2 is the diffraction angle of the strongest
peak.

Table 1
Compositions of WO3-epoxy composites with different volume fractions of ller WO3 and
epoxy resin.
Composite by volume fraction (vol%)
Filler (WO3)

Epoxy resin

2
4
6
8
10

98
96
94
92
90

Fig. 1. Comparisons of X-ray transmission comparison in nano-sized and micro-sized

WO3-epoxy composites for synchrotron radiation generated by the XAS unit for energy
of 10 40 keV.

4954

N.Z. Noor Azman et al. / Materials Science and Engineering C 33 (2013) 49524957

Fig. 2. The X-ray transmission ratio of micro-sized WO3-epoxy composite (I/Io)m to

the X-ray transmission for nano-sized WO3-epoxy composite (I/Io)n, (I/Io)m/(I/Io)n
for synchrotron energies (1040 keV).

unit generates characteristic X-ray energies of molybdenum (17.5 keV

and 19.6 keV) or rhodium (20.2 keV and 22.7 keV), it is much easier
to compare with the XAS results (see Fig. 3). In the results presented
in Fig. 3, it is clearly shown that the X-ray transmission results for the
mammography unit sat between the results of 1525 keV for the XAS
beam energies. In contrast, for the radiography unit, the operated X-ray
tube voltages generated a broad spectrum (polychromatic X-ray beam).
Thus, the equivalent energies for the X-ray tube voltages of the radiography unit were estimated from the XAS results by superimposing their
data together (see Fig. 4). As can be seen in Fig. 4, the X-ray transmissions
of samples for X-ray tube voltages of 4060 kV were sitting between 25
and 40 keV while the others were sitting above 40 keV. Hence, the Xray tube voltages of 4060 kV operated by the radiography unit produced
the equivalent X-ray energies in the range of 2540 keV while the X-ray
tube voltages of 60 kV had an equivalent energy of 40 keV.
Fig. 5 shows the comparison of X-ray transmissions for samples with
different thicknesses. Samples of 7 mm thickness were used in our previous study [19], whereas 2-mm-thick samples were used in this study.
Fig. 5a and b provide the results for the mammography unit only on
samples with 4 vol% and 6 vol% loading of WO3, respectively. As shown
in these gures, the differences in the X-ray transmission between the
micro-sized and nano-sized WO3-epoxy became larger for thicker samples (7 mm). In contrast, Fig. 5c shows insignicant differences in X-ray
transmission between nano-sized and micro-sized WO3-epoxy composites for radiography tube voltages only for samples with a loading
of 4 vol% WO3. This trend was also observed for all the other loadings
(i.e., 2, 6, 8 and 10 vol% WO3). These ndings are in good agreement
with the work by Knzel and Okuno (2012), which also showed that

Fig. 3. Comparisons of X-ray transmission in nano-sized and micro-sized WO3-epoxy composites for synchrotron radiation generated by the XAS unit for energies of 1025 keV and
mammography unit tube voltages of 2549 kV.

Fig. 4. Comparisons of X-ray transmission in nano-sized and micro-sized WO3-epoxy composites for synchrotron radiation generated by the XAS unit for energies of 2540 keV and
radiography unit tube voltages of 40100 kV.

the grain size effect increased with the increase of the sample thickness
at low energy X-ray beams (25 kV and 30 kV) but remained unchanged
over the material thickness for higher energy X-ray beams (60 kV) [18].
In general, the photoelectric effect is the most likely interaction to
occur within a matter at a lower photon (X-ray) energy range. In this interaction, a photon will transfer its entire energy to an electron in the
material on which it impinges. The electron thereby acquires enough
energies to free itself from the material to which it is bound and then
may undergo single or multiple-scattering events with neighboring
atoms. In addition, there is also a slight uctuation in the probability
of emission of Auger electrons and uorescent photons may form during this interaction. This phenomenon can contribute to the alteration
of the mass attenuation coefcient of an element relative to the bulk
material when considered over a small range of X-ray energies. The
probability of photoelectric interaction is directly proportional to the
cube of the atomic number of the absorbing material Z3 and inversely
proportional to the cube of the X-ray energy (1/E)3.
Moreover, nano-sized WO3-epoxy composites consist of a higher
number of WO3 particles/gram when compared to micro-sized WO3epoxy composites. Therefore, the distribution of the nano-sized WO3
in the resin should also be different from that presented by microsized WO3, thus resulting in a more uniform dispersion in the resin. As
a consequence, the chances of an X-ray photon with lower energy to interact and be absorbed by WO3 particles may be higher in nano-sized
WO3-epoxy composites than in micro-sized WO3-epoxy composites.
Fig. 6 shows the back-scattered images of the same loading of WO3
(4 vol%) within nano-sized WO3-epoxy and micro-sized WO3-epoxy
composites using the Zeiss Evo 40XVP scanning electron microscope.
The WO3 particles were seen to be more closely dispersed in the
nano-sized WO3-epoxy composite (Fig. 6a) as compared to its microsized counterpart (Fig. 6b). Thus, the probability for the lower energy
photons to interact with the WO3 particles and be absorbed is higher
for the nano-sized WO3-epoxy composite.
As the photon energy increases, the photon (X-ray) penetration
through the absorbing material without interaction increases, and
hence, less photoelectric effect relative to the Compton effect occurs.
Thus, the X-ray attenuation by the absorbing material decreased since
the Compton interaction was weakly dependent on Z and E and this
interaction only took place between the incident photon and one of
the outer shell electrons of an atom in the absorbing material.
In order to discover the X-ray shielding ability of the composites, the
results were compared to commercial lead (Pb) sheets (model RAS20
Calibrated Absorber Set) of four different thicknesses (i.e., 0.81, 1.63,
3.18 and 6.35 mm) (see Fig. 7) using a radiography unit of tube voltages
40100 kV. The results show that although the lead sheets gave the
lowest X-ray transmissions at each tube voltage when compared to all
compositions of WO3-epoxy composites, the latter with 10 vol% of
either micro-sized or nano-sized WO3, can be a substitute for Pb in

N.Z. Noor Azman et al. / Materials Science and Engineering C 33 (2013) 49524957

4955

Fig. 6. SEM images for epoxy composites lled with (a) 4 vol% nano-sized WO3 and (b)
4 vol% of micro-sized WO3.

all of these databases was chosen to be the same as the wavelength of

the synchrotron radiation used. The diffraction peaks shown in Fig. 8
belong to monoclinic WO3 (PDF le 00-043-1035). These results indicate that both the micro-sized WO3 and the nano-sized WO3 were
single-phase pure without impurities.

Fig. 5. X-ray transmission comparison for different thickness of the sample for (a) 4 vol%
of WO3 ller epoxy composites for mammography unit tube voltages of 30, 35 and 49 kV;
(b) 6 vol% of WO3 ller epoxy composites for mammography unit tube voltages of 30, 35
and 49 kV; and (c) 4 vol% of WO3 ller epoxy composites for radiography unit tube voltages of 40, 60, 80 and 100 kV.

X-ray shielding by increasing the sample thickness to 7 mm. Hence,

the usage of lead in X-ray shielding can be substituted by WO3 whereby
W is lighter and less toxic compared to Pb.
3.2. Crystallite size
The reference for tting the peaks was taken from the International
Centre for Diffraction Data PDF-4+ 2009 database. The wavelength for

Fig. 7. X-ray transmission as a function of radiography unit X-ray tube voltage

(40100 kV) for all micro-sized and nano-sized WO3-epoxy composites and commercial
lead sheets.

4956

N.Z. Noor Azman et al. / Materials Science and Engineering C 33 (2013) 49524957

The diffraction patterns were plotted only in the 2 range of 915

to clearly show the size difference of the peaks for each micro-sized and
nano-sized WO3. As shown in Fig. 8, the broad peaks belonged to
nanometer-sized WO3 crystallite whereas the well-dened crystalline
peaks belonged to micrometer-sized WO3. The crystallite size determined from the Scherrer equation for nano-sized WO3 was 51.5 nm.
These nano-crystallites were signicantly smaller than the particle
sizes provided by Sigma-Aldrich, thus indicating that at least 2 crystallites were present in each WO3 particle of 100 nm in size.
3.3. Mechanical properties
Fig. 9 shows the effect of ller loading on the exural strength,
exural modulus and Rockwell hardness of the epoxy composites. The
exural strength was found to decrease with increased WO3 ller
size while it had little or no effect on exural modulus. Similar results
were reported by Park [21], who observed an increase in exural
strength with decreased particle size in silica-reinforced epoxy composites. However, Moloney et al. [22] reported a negligible effect of
particle size on exural modulus in their epoxy composites lled
with silica.
From Fig. 9a, it can be seen that the exural strength of pure epoxy
was 49.9 MPa, but increased to a maximum value of 64 MPa for the
composite containing 4 vol% nano-sized WO3. However, a further
increase in the ller loading beyond 4 vol% resulted in a decrease
in exural strength whereby the composite containing 10 vol%
nano-sized WO3 exhibited the lowest exural strength of 52.6 MPa.
Similarly for micro-sized WO3-epoxy composites, the maximum
exural strength was obtained for a ller loading of 2 vol%. A reduction in exural strength was again observed when the ller loading
was increased beyond 2 vol% due to non-uniform dispersion of the
ller within the matrix. The resultant agglomeration of the llers
acted as stress-concentrators which served to reduce the strength
of the composites.
The exural modulus of the composites increased with an increase
in the ller loading for both the nano-sized and micro-sized WO3,
which may indicate that the stiffness of these composites obeyed the
well-known rule-of-mixtures (Fig. 9b).
Finally, the hardness results presented in Fig. 9c indicate that an
increase in the WO3 ller loading resulted in an initial increase in the
hardness of the composite, but a further increase in ller loading at
10 vol% caused a reduction in hardness probably due to the undesirable
agglomeration of the llers. The initial increase in the hardness of the
composites observed for both the nano-llers and the micro-llers
may be attributed to their uniform dispersion within the epoxy matrix,

Fig. 9. Mechanical properties of epoxy composites showing: (a) exural strength as a function of WO3 ller loading; (b) exural modulus as a function of WO3 ller loading; and (c)
Rockwell hardness as a function of WO3 ller loading.

together with their strong interaction with the epoxy chains to form
good interfacial bonding.
4. Conclusion

Fig. 8. Typical powder diffraction patterns for micro-sized and nano-sized WO3 loading in
epoxy composites.

The size effect of WO3 particles on the X-ray transmission in nanosized and micro-sized WO3-epoxy composites has been investigated
at various synchrotron radiation energies (i.e., 1040 keV). The results

N.Z. Noor Azman et al. / Materials Science and Engineering C 33 (2013) 49524957

presented in this work demonstrated that the size effect on X-ray attenuation was profoundly dependent on the energy of the synchrotron
radiations. The particle size effect was more pronounced at lower synchrotron radiation energies (1020 keV) since the X-ray transmission
in nano-sized WO3-epoxy composites was less than in their microsized counterparts. However, this size effect became insignicant at
higher energies of 2040 keV because the X-ray transmissions in both
nano-sized and micro-sized WO3-epoxy composites were very similar.
The X-ray transmission results for the mammography unit sat between
the results of 1525 keV for XAS beam energies. Meanwhile, the X-ray
transmissions in samples for X-ray tube voltages of 4060 kV of the radiography unit sat between 25 and 40 keV. In addition, for composites
with the same ller loading, but with increasing sample thickness, the
size effect in X-ray transmission was most prominent for X-ray tube
voltages of 2535 kV but was negligible at 35100 kV. As the ller loading of the WO3 increased, the mechanical properties showed an initial
optimum improvement, but a further increase in the ller loading
caused these properties to deteriorate.

Acknowledgements
The collection of X-ray absorption spectroscopy (XAS) data was
funded by the Australian Synchrotron (AS123/XAS5341). We thank
Dr. Bernt Johannessen of the Australian Synchrotron and our colleagues
Dr. C. Ng and A/Prof. Z. Sun for assistance with XAS data collection. Also,
we would like to thank Carolyn Madeley of Breast Assessment Centre,

4957

Royal Perth Hospital, Western Australia for giving us the opportunity

to use the mammography unit.
References
[1] C.R. Patra, R. Bhattacharya, D. Mukhopadhyay, P. Mukherjee, Adv. Drug Deliv. Rev.
62 (2010) 346361.
[2] F. Van Den Heuvel, J.P. Locquet, S. Nuyts, Phys. Med. Biol. 55 (2010) 45094520.
[3] T. Wang, Z. Liu, M. Lu, B. Wen, Q. Ouyang, Y. Chen, C. Zhu, P. Gao, C. Li, M. Cao, L. Qi, J.
Appl. Phys. 113 (2013) 024314024318.
[4] B. Hoyska, Spectrochim. Acta B At. Spectrosc. 24 (1969) 8593.
[5] A. Granmayeh Rad, H. Abbasi, M.H. Afzali, Phys. Procedia 22 (2011) 203208.
[6] F. El Haber, G. Froyer, J. Univ. Chem. Technol. Metall. 43 (2008) 283290.
[7] P. Jackson, S. Periasamy, V. Bansal, M. Geso, Australas. Phys. Eng. Sci. Med. 34 (2011)
243249.
[8] X. Huang, M.A. El-Sayed, J. Adv. Res. 1 (2010) 1328.
[9] P.D. Sahare, R. Ranju, S. Numan, S.P. Lochab, J. Phys. D Appl. Phys. 40 (2007) 759.
[10] A. Popov, SPIE Newsroom 24 (2009) 12.
[11] S.-S. Chen, H.-C. Chen, W.-C. Wang, C.-Y. Lee, I.N. Lin, J. Guo, C.-L. Chang, J. Appl. Phys.
113 (2013) 113704113710.
[12] M. Steinhart, Angew. Chem. Int. Ed. 43 (2004) 21962197.
[13] M.Z. Botelho, R. Knzel, E. Okuno, R.S. Levenhagen, T. Basegio, C.P. Bergmann, Appl.
Radiat. Isot. 69 (2011) 527530.
[14] Lablogic, in: L.S. Limited (Ed.), Lablogic, Shefeld, United Kingdom, 2009.
[15] S. Daren, POLYMICRO Newsletter, 2004.
[16] Premac Lead Acrylic, Part of the Wardray Premise Total Radiation Shielding Package,
Wardray Premise Ltd., Surrey, United Kingdom, 2005-2011.
[17] M. Faccini, C. Vaquero, D. Amantia, J. Nanomater. 2012 (2012) 19.
[18] R. Knzel, E. Okuno, Appl. Radiat. Isot. 70 (2012) 781784.
[19] N.Z. Noor Azman, S.A. Siddiqui, R. Hart, I.M. Low, Appl. Radiat. Isot. 71 (2013)
6267.
[20] Annual Book of ASTM Standards, vol. 08.01, ASTM, 2005.
[21] J.J. Park, Trans. Electr. Electron. Mater. 14 (2013) 3942.
[22] A.C. Moloney, H.H. Kausch, T. Kaiser, H.R. Beer, J. Mater. Sci. 22 (1987) 381393.